Neutralizing competition in CRISPRi genetic circuits

In CRISPRi-based circuits, dCas9 is a shared resource among multiple repressors. To ensure that multiple repressors can operate concurrently without interfering with one another, we designed a dCas9 concentration regulator.
Neutralizing competition in CRISPRi genetic circuits

Thanks to the introduction of CRISPR as a routine gene editing tool, biotechnologies are living a new age of discoveries and advancements boosted by the high programmability and precision of this molecular platform. CRISPR technology has gone long way over its classic use for “cutting DNA”, and, in particular, has been employed also for gene expression regulation. Synthetic biology has adopted with great excitement this technology with an aim of creating large libraries of orthogonal regulators of gene expression. In fact, a large set of regulators that do not interfere with one another could allow to scale up the size and sophistication of synthetic genetic circuits, which has been limited by the scarcity of transcriptional regulators. I was very curious about the reason for this excitement.  To what extent CRISPRi had allowed to finally have a large number of regulators work concurrently in the cell without interfering with one another? In our recent Nature Communication paper, we show that multiple CRISPRi-based, orthogonal, regulators that work concurrently in the cell actually interfere with one another due to competition for dCas9 and we engineer a solution.

The story behind this paper started in the Fall of 2016, when I decided to explore a mathematical model of CRISPR-based genetic circuits to understand where the key enabler to scalability was. With Penny Chen, a first-year graduate student at the time, we constructed and analyzed a mathematical model of genetic circuits based on CRISPR interference (CIRISPRi). In CRISPRi, catalytically inactive Cas9 protein (dCas9) is recruited by a small guide RNA (sgRNA) to a prescribed target promoter and impedes RNA polymerase’s binding. The idea that supports scalability is that one can design many different sgRNAs to recruit dCas9 to any target promoter with very high specificity.   But how can high binding specificity alone rule out interference among multiple sgRNAs? Indeed, the model revealed that multiple sgRNAs’ regulatory paths interfere with one another despite highly specific binding of any sgRNAs to its prescribed target. This interference occurs because sgRNAs compete for dCas9, which is a shared and limited resource in CRISPRi-based genetic circuits. As a consequence, expression of one sgRNA elicits an indirect effect on the targets of any other sgRNA by sequestering dCas9, as we documented in our early modeling paper [1]. Thus, efforts spent to design multiple sgRNAs such that their bindings are orthogonal to one another are not serving their purpose unless dCas9 competition among sgRNAs is removed. During late 2017, Massimo Bellato, at the time a PhD student at the University of Pavia (Italy), joined our group as a visiting researcher to work on problems of ribosome sharing in genetic circuits in collaboration with other students in my group. It was by chance that Massimo had been working on CRISPRi for his PhD thesis and that he already had available some of the genetic constructs that we wanted to build to validate our model. This was good timing! With Massimo’s lead, we quickly validated experimentally our model predictions. We demonstrated that expression of one sgRNA not only leads to repression of its target promoter but, due to dCas9 competition, also to upregulation of promoters targeted by different sgRNAs. 

While on the one hand we were excited to see that the experiments confirmed our mathematical models, on the other hand we were also disappointed to find out that the expectations for scalability were not met by CRISPRi-based genetic circuits. At this point, I was working in my lab on the implementation of a quasi-integral controller to mitigate the coupling among genetic modules due to competition for ribosomes, which we later published in Nature Communications [2]. Therefore, I decided to exploit a similar feedback control approache to design a genetic feedback regulator to keep the level of dCas9 in CRISPRi circuits at a constant level, independent of variable demand by sgRNAs. This effort was led by research scientist Hsin-Ho Huang in my lab. Essentially, the feedback regulator increases the level of total dCas9 when more sgRNAs are expressed and decreases total dCas9 concentration when less sgRNAs are expressed. In doing so, the amount of dCas9 available to any sgRNA stays constant independent of other sgRNAs’ expression, such that interference between any two sgRNAs is removed.  This effectively allows independent operation of multiple CRISPRi-based regulators concurrently in the cell. One aspect of the regulator is also that it keeps the concentration of free dCas9 at a sufficiently low level to avoid toxicity. To achieve this, while allowing the free dCas9 concentration to be unperturbed by the expression of sgRNAs, the regulator combines a strong promoter expressing dCas9 with a strong repression of dCas9 through CRISPRi. This approach is called high-gain negative feedback in control systems. High-gain negative feedback is at the basis of early designs in electronics that allowed to reject loads upon composition of circuit modules, hence enabling modularity and scalability. 

Our solution suggests that, just like feedback control has been instrumental to achieve modularity and scalability in electronic circuit design, it will likely play a similar critical role in enabling modular and scalable design in engineering biology. Overall, we believe that the dCas9 concentration regulator, by neutralizing competition, will allow true scalability of CRISPRi-based genetic circuits wherein many regulators can operate concurrently in the cell without interference.


[1] Chen, P.-Y., Qian, Y. & Del Vecchio, D. A Model for Resource Competition in CRISPR-Mediated Gene Repression. IEEE Conference on Decision and Control (CDC) 4333–4338 (2018).

[2] Huang, H.-H., Qian, Y. & Del Vecchio, D. A quasi-integral controller for adaptation of genetic modules to variable ribosome demand. Nat. Commun. 9:5415 (2018).

Please sign in or register for FREE

If you are a registered user on Nature Portfolio Bioengineering Community, please sign in